Auditable Emergency Triage for Maternal and Newborn Care in India

印度母婴护理的可审计紧急分诊系统

Abstract: At Noora Health, our nurses answer more than 50,000 medical queries per month on our WhatsApp-based service that provides caregivers with on-demand support. Their most time-critical task is emergency triage: deciding which queries need immediate in-person attention. To support them, we built a system that uses a large language model (LLM) to classify whether a message is an emergency and provide a rationale for interpretability. But the system was opaque: analyzing mistakes meant reading reasoning chains for each message, which is infeasible at our scale. Prompt changes meant re-running a full evaluation to prevent regressions, which was both costly and operationally challenging. Clinicians follow a decision tree to make this call, but it was never documented or passed to the model, which relied on a flat list of danger signs.

摘要： 在 Noora Health，我们的护士每月通过基于 WhatsApp 的服务回答超过 50,000 条医疗咨询，为护理人员提供按需支持。她们最紧迫的任务是紧急分诊：决定哪些咨询需要立即进行线下就诊。为了提供支持，我们构建了一个系统，利用大语言模型 (LLM) 来判断消息是否属于紧急情况，并提供可解释的理由。然而，该系统是不透明的：分析错误意味着必须阅读每条消息的推理链，这在我们的规模下是不可行的。修改提示词意味着需要重新进行全面评估以防止性能倒退，这既昂贵又在操作上极具挑战性。临床医生通常遵循决策树来做出判断，但这些决策树从未被记录或传递给模型，模型仅依赖于一份扁平的危险信号列表。

To address these issues, we decomposed triage into two steps: an LLM extracts canonical symptoms and patient context from the query using a clinician-authored vocabulary, and a deterministic rule engine captures the scenarios that indicate an emergency. We show that the new system raised recall from 0.565 to 0.810 and F1 from 0.606 to 0.702, with structured rules driving most of the accuracy gains while the decomposition provides auditability: clinical experts can inspect each stage of the new system to see whether the query was mistranslated, symptoms were incorrectly extracted, patient context was wrongly inferred, or the necessary rules were missing. They can add new rules independently without causing regressions and avoid running costly evaluations.

为了解决这些问题，我们将分诊过程分解为两个步骤：首先，LLM 使用临床医生编写的词汇表从咨询中提取规范化的症状和患者背景；其次，确定性规则引擎捕捉指示紧急情况的场景。研究表明，新系统将召回率从 0.565 提高到 0.810，F1 分数从 0.606 提高到 0.702。结构化规则推动了大部分准确率的提升，而分解过程则提供了可审计性：临床专家可以检查新系统的每个阶段，查看咨询是否被误译、症状提取是否错误、患者背景推断是否偏差，或者是否缺少必要的规则。他们可以独立添加新规则，而不会导致性能倒退，也避免了进行昂贵的评估。

Since deployment, the new system has triaged 152,421 patient queries and flagged 28,535 (18.7%) as emergencies. The over-escalation rate has been 17.8%, without any increase in missed emergencies. Clinicians have also added 48 new rules since deployment, evidence of the faster correction loop we set out to build.

自部署以来，新系统已分诊了 152,421 条患者咨询，并将其中 28,535 条 (18.7%) 标记为紧急情况。过度升级率 (over-escalation rate) 为 17.8%，且未出现漏报紧急情况的情况。自部署以来，临床医生还添加了 48 条新规则，这证明了我们旨在构建的更快速的纠错循环已初见成效。